Delimit a vector by decimal

Hi,

After importing my data and delimiting by white space, I am left with two columns. For example, I am left with a matrix with a column of

1.2340 or similar numbers like 122.433

and a column of

233 or similar numbers

My problem is that the left column has to be split at the decimal into two separate numbers--but exactly how it appears. If the number on the right hand side of the decimal ends in a 0, I need to preserve it, as this isn't actually a decimal, but two identifying numbers stuck together with a decimal. Every solution I have tried thus far that has separated the numbers have dropped the 0. Any suggestions?

[Azzi Abdelmalek]

Not clear

[Mason]

Ok, to clarify, just assume that I have a column of numbers such as

1.233

22.330

540.3001

What I need to do here is separate this column into two separate columns; one containing the digits on the left of the decimal, and the other containing the digits on the right of the decimal.

My problem so far has been keeping the 0 in the second entry; everything I have tried has shortened 330 to 33.

[Azzi Abdelmalek]

Have you imported those data as a char or as double?

[Mason]

double

[Azzi Abdelmalek]

Ok, but why one number is with 3 decimals and another with 4 decimals? It's not possible

[Azzi Abdelmalek]

Check this

A=[1.233;22.330;540.3001]

The result

A =

    1.2330
   22.3300
  540.3001

[Azzi Abdelmalek]

It's better if you post a sample of your txt file

[Mason]

I'm not sure why, but when I import the data that I am working with, it imports it fine.

The only thing I'm having a problem with is separating them.

[Azzi Abdelmalek]

What is the difference between 1.30 and 1.300?

[dpb]

But you'll play h, e, double-hockey sticks separating them as doubles.

Import as cell array of strings then split on the character decimal, then convert those substrings to numeric. That way you'll keep the '330' example as the full substring to convert instead of just the numeric remainder.

Or, of course, fix the broken data encoding scheme in the generating process and export two columns of properly scaled values instead of the combination.

[Mason]

Here is a sample:

3001.1 0.082086927956593

3001.2 0.000152538100647958

3001.3 4.77296878570862E-05

3001.11 0.000174921404962367

3001.12 0.000147929354099336

3001.65 0.0254316164029534

3001.120 0.947370074332119

3001.121 0.0215693042587116

3001.122 0.00378786288621699

3001.126 0.382907657356411

3001.132 0.00166090942239532

3001.133 0.175693456921313

the numbers on the right are irrelevant; they get their own column when I import the data and delimit by white space. They are actually decimal values.

The numbers on the left, for example

3001.1

are the ones I need to separate.

This becomes a problem at the entry

3001.120

When I try to separate this column by the decimal, it will drop the 0, and the 120 will become 12.

[Azzi Abdelmalek]

If you import 1.3 and 1.300 as double, the result is the same

[Mason]

I'm not sure how I was successfully getting the different numbers, then...but do you have a way around this?

[Mason]

@dpb, the broken data encoding scheme is actually that of an outside company...so frustrating! Haha

But, how would I import a DAT file as a cell array? I'm not familiar with the process, though it is probably simple?

Answer 1

[Azzi Abdelmalek]

If you import your data as cell array of char

A={'1.233';'22.330';'540.3001'}
out=regexp(A,'\.','split')
out=str2double(reshape([out{:}],2,[])')

[Mason]

This definitely works when I manually input A.

But how do I get A to be an imported DAT file?

[Azzi Abdelmalek]

fid=fopen('file.dat')
M=textscan(fid,'%s %s')
fclose(fid)
A=M{1}

[Mason]

Thank you!